You don't have javascript enabled.

Loan Portfolio Optimisation: Early Warning is the key

According to a recent European Banking Authority report, overall risk indicators and market sentiment are strengthening, however, the signs of recovery remain modest and fragile. The quality of loan portfolios continue to decline which remains a concern. Credit quality was, and still is, the main source of risk. As a result, credit monitoring has become

  • Nicole Miskelly
  • January 9, 2015
  • 4 minutes

According to a recent European Banking Authority report, overall risk indicators and market sentiment are strengthening, however, the signs of recovery remain modest and fragile. The quality of loan portfolios continue to decline which remains a concern. Credit quality was, and still is, the main source of risk. As a result, credit monitoring has become a priority for financial institutions, and Early Warning systems are becoming essential for improving portfolio performance. Davide Capuzzo, Analytics Director of CRIF gives bobsguide his insights.

What is the role of these systems within credit institutions today?

The role of Early Warning systems is to predict the deterioration of credit positions as early as possible, enabling managers to find solutions which avoid the borrower defaulting, or in any case limit the economic impact. Nowadays, their role is changing from a simple aid for managers to manage loan portfolio assets, to a system based not only on priorities but also on well-defined rules and responsibilities, which are in turn monitored. The subject of Early Warning systems is a hot topic because, whereas it previously meant 'identification', today it is closely associated with ‘action’ ‘and results.

But is this not the job of internal rating systems?

Internal rating systems are an integral part of Early Warning systems, but do not and cannot perform this function on their own for a number of reasons. They are in fact developed to give stable risk assessments over time at the expense of the reactivity required of an Early Warning system. They are not fed with the same types of event which characterise those systems, which include daily events. The situation predicted by an internal rating system is typically loan default, whereas Early Warning systems predict a status well before default, or even the simple deterioration of the status. Therefore, from a modeling point of view, they are two integrated systems, but which serve different purposes and processes.

What are the key features of an Early Warning system?

These can be summed up in three characteristics. The first is that they are complementary to internal rating systems in terms of the timing and variety of the indicators used to predict the deterioration. In particular, often information is acquired which is not always used in internal assessment processes, including, for example, credit bureau and business information. The second relates to the design of a system of actions and responsibilities managed by monitoring and customer management. The third aspect comprises a process of cultural change among staff and progressive fine-tuning of the management objectives within the various teams.

How are these monitoring systems likely to develop in the future?

On the one hand, there will be a widening range of information considered, particularly internal data, which in the past were not managed electronically, and external information such as credit information and alternative and “big” data sources. On the other hand, there will also be a refinement of Early Warning processes using the measurement systems and controls mentioned before. This type of development will be based on:

Advanced management systems characterized by the ability to integrate with non-traditional information sources, to implement complex management strategies for high-risk positions, and to produce reporting with the measurement of manager performance.

The existence of a laboratory-type environment, where basic information, indicators, proposed actions, timing and methods of application, and quantitative results are stored for the purposes of monitoring and measuring the performance of the monitoring system. The Early Warning models and related actions can then be adjusted, if required. This last aspect is particularly significant in light of recent central bank regulatory provisions on the control of risk management monitoring systems.

In reality, this is not completely new and, in this regard, it is expected that over time Early Warning systems will incorporate some approaches from collection systems, which in some credit sectors are already characterised by significant industrialisation and efficiency.